Healthy Donor Single Cell Cytof Analysis Workflow

1) viability gating 2) Minibatch K means clustering 3) Annotating downsampled clusters w/ clustergrammer 4) Map annotations back to single cell 5) Generate figures: population frequencies and final UMAP

For some reason, there are more cells in 'meta_cell', than 'cell', so we'll drop meta_cells that do not exist in cell. We can then verify that the 'cell' and 'meta_cell' indices match.

Arcsinh transform CyTOF data

Viability Cleanup

Here, we'll utilize a bokeh script written by Darwin to manually gate on Rh103-Pt195- Live Cells.

Mini Batch K means downsampling of data

Manual Annotation of Downsampled Clusters with Clustergrammer

Here, we will iteratively annotate the downsampled clusters in a tierwise fashion, increasing in granularity by tier

Manual annotation Tier 2

For each Tier1 annotation, we will load a heatmap with just those clusters (i.e. only T cells, and annotate those T cells more granulary and save to Tier2.

Tier 2 annotations complete, let's view final heatmap

Map annotations back to single cell

Figures